In a SS7 and/or SIGTRAN system, congestion and Send Failure Notification have been implemented. Congestion or send failure appears when the peer do not acknowledge messages fast enough compared to the traffic rate in the interface between the peer node and the layer 2 instance on the local node. This means that in case the traffic rate is too high compared to the throughput of the interface between peer's layer 2 instances, a congestion indication or a send failure notification will be sent by the local layer 2 instance to the local layer 3 instance to indicate that there is an congestion situation, or overload situation. This indication is then forwarded to the client on top of layer 3 or forwarded to other nodes in case the local node acts as a gateway.
The mechanism to detect an overload situation in the interface between layer 2 and layer 3 for the incoming direction is known from the above mentioned systems.
Layer 3, for example M3UA, buffers data when it receives a Send Failure Notification (SFN) in existing implementations of SS7 network protocol stacks.
Presently, flow control for outgoing SS7 traffic is handled by the use of process priorities. The clients have had lower priority, thereby making it impossible to load the interface between layer 2 and layer 3 higher than the processing unit can manage.
As a result of the introduction of multi core processors in CPP (Connectivity Packet Platform), present solutions for flow control need to be improved. The process priorities are effectively disabled by the multi core environment. Due to this there is a need to implement a flow control function to make sure that the system is not overloaded by messages. The overload can cause various problems like lost data or crashing processes.